AI Guides › Career & Work

Using AI To Prepare For A Performance Review, Both Sides Of It

By Nigel Guy · 2 min read

Most performance-review prep happens the night before, from memory, which means it's built on whatever you happen to remember rather than what actually happened over the past six or twelve months. Managers do the same thing on their side. Both walk in less prepared than they think.

The rule: prepare from a real record, not from memory — and prepare both the case you're making and the pushback you'll get, not just the highlight reel.

The mechanism

If you're the one being reviewed:

  1. Pull your actual record, not your recollection of it — completed projects, messages, tickets, whatever exists. Memory over-weights recent events and under-weights the quiet, steady work from months ago.
  2. Draft your own summary from that record, organised by impact, not by chronology. What changed because of your work, specifically.
  3. Ask for the honest counter-argument. What would a sceptical manager say is missing or overstated in this draft? Answering that in advance beats hearing it live for the first time.
  4. Prepare one real ask, not a vague one — a specific project, a specific scope change, a specific number, with the reasoning attached.

If you're the one giving the review:

  1. Do the same record-pull for the person you're reviewing — recency bias affects managers as much as anyone.
  2. Draft specific, checkable feedback, not general impressions. "Missed two deadlines in Q2, both flagged late" is checkable; "needs to manage time better" is not.
  3. Prepare the harder feedback first, in draft, so it isn't softened or avoided live out of discomfort in the moment.

What to skip

Skip using a draft summary as your literal script — read it, absorb it, speak in your own words. And skip inflating the draft with achievements that aren't really yours; a review built on an overstated record falls apart under one honest follow-up question.

Guardrails

All 751 AI guides · JulieMango plans from £17/mo